• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.13-14
• /
• 2023

집중 안전 점검의 대상인 노후 건축물에서 균열은 건물의 안전도를 점검할 수 있는 지표이다. 안전 점검에 드론을 활용하면서 고해상도의 드론 기반 균열 이미지 수집이 가능해졌고, 육안이 아닌 AI기반으로 균열을 탐지, 구획화할 수 있다. 본 연구에서는 주변 사물과 배경에 구애받지 않고 안전 점검이 가능한 구획화 알고리즘을 제안한다. METU와 POC데이터셋을 가공하여 데이터셋을 구축하고, 이를 바탕으로 ResNet50을 통해 균열과 유사한 배경을 분류하였으며, 균열 구획화 모델을 선정하여 DesneNet201-UNet++으로 mIoU 82.27%를 달성하였다. 본 연구는 노후 건축물 안전 점검에 필요한 균열 폭 추정에 도움이 될 것으로 기대된다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.7 no.5
• /
• pp.934-940
• /
• 2003

In this paper, hair beauty fashion design feature points detection system is proposed. A hair models and hair face is represented as a graph where the nodes are placed at facial feature points labeled by their Gabor features and the edges are describes their spatial relations. An innovative flexible feature matching is proposed to perform features correspondence between hair models and the input image. This matching hair model works like random diffusion process in the image space by employing the locally competitive and globally corporative mechanism. The system works nicely on the face images under complicated background. pose variations and distorted by accessories. We demonstrate the benefits of our approach by its implementation on the face identification system.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2024.01a
• /
• pp.347-348
• /
• 2024

본 논문에서는 철도 구성요소 모니터링을 위한 효율적인 객체 분할 기법으로 사전학습된 SegFormer 모델의 적용을 제안하고, 객체 분할을 위해 보편적으로 사용되는 U-Net 모델과의 성능 비교 분석을 진행하였다. 철도의 주요 구성요소인 선로, 침목, 고정 장치, 배경을 분할할 수 있도록 라벨링된 데이터셋을 학습에 사용하였다. SegFormer 모델이 대조군인 U-Net보다 성능이 Jaccard Score 기준 5.29% 향상됨에 따라 Vision Transformer 기반의 모델이 기존 CNN 기반 모델의 이미지의 전역적인 문맥을 파악하기 상대적으로 어렵다는 한계를 극복하고, 철도 구성요소 객체 분할에 더욱 효율적인 모델임을 확인한다.

• Journal of KIISE:Software and Applications
• /
• v.37 no.9
• /
• pp.712-721
• /
• 2010

This paper proposes a new fire-flame detection method using probabilistic membership function of visual features and Fuzzy Finite Automata (FFA). First, moving regions are detected by analyzing the background subtraction and candidate flame regions then identified by applying flame color models. Since flame regions generally have continuous and an irregular pattern continuously, membership functions of variance of intensity, wavelet energy and motion orientation are generated and applied to FFA. Since FFA combines the capabilities of automata with fuzzy logic, it not only provides a systemic approach to handle uncertainty in computational systems, but also can handle continuous spaces. The proposed algorithm is successfully applied to various fire videos and shows a better detection performance when compared with other methods.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.1
• /
• pp.1-8
• /
• 2010

Some methods employing the Active Contour Model have been widely used to extract face contour. Their performance, however, depends on the initial position of the model and the coefficients of the energy function which should be reconsidered whenever illumination and environmental condition of an input image is changed. Additionally, the number of points in the contour model should increase drastically in order to extract a fine contour. In this paper, we thus propose a novel approach which extracts face contour by segmenting the face region with threshold values obtained by a histogram analysis technique in the separated region of input image. The proposed method shows good performance under various illumination and environmental condition since it extracts face contour by considering the characteristics of the input image.

• Journal of the Institute of Electronics Engineers of Korea TE
• /
• v.42 no.1
• /
• pp.61-68
• /
• 2005

Image-based rendering is an approach to generate realistic images in real-time without modeling explicit 3D geometry. Especially, owing to its simplicity, TIP(Tour Into the Picture) is preferred to constructing a 3D background scene. Because existing TIP methods have a limitation in that they lack geometrical information, we can not expect a accurate scene if the viewpoint is far from the origin of the TIP. In this paper, we propose the method of constructing a virtual environment of a wide area by blending multiple TIP images. Firstly, we construct multiple TIP models of the virtual environment. Then we interpolate foreground and background objects respectively, to generate a smooth navigation image. The method proposed here can be applied to various industry applications, such as computer game, 3D car navigation, and so on.

• Proceedings of the Korea Contents Association Conference
• /
• 2005.11a
• /
• pp.633-638
• /
• 2005

The extraction of a face is a very important part for human interface, biometrics and security. In this paper, we applies DCM(Dilation of Color and Motion) filter and Active Contour Models to extract facial outline. First, DCM filter is made by applying morphology dilation to the combination of facial color image and differential image applied by dilation previously. This filter is used to remove complex background and to detect facial outline. Because Active Contour Models receive a large effect according to initial curves, we calculate rotational degree using geometric ratio of face, eyes and mouth. We use edgeness and intensity as an image energy, in order to extract outline in the area of weak edge. We acquire various head-pose images with both eyes from five persons in inner space with complex background. As an experimental result with total 125 images gathered by 25 per person, it shows that average extraction rate of facial outline is 98.1% and average processing time is 0.2sec.

• Journal of the Institute of Convergence Signal Processing
• /
• v.13 no.1
• /
• pp.44-49
• /
• 2012

The indirect-contact ECG measurement is a newly developed method for unconstrained and nonconscious measurement in daily Life. This study is the first step to reducing the large background noise appearing in indirect-contact ECG. This study built the thermal noise model of capacitively coupled active electrode which is used in indirect-contact ECG. The results show that the level of thermal noise estimated by the thermal noise model is much the same as that of actual background noise for the capacitively coupled active electrode alone. By applying the actual electrical properties of a sample cotton cloth to the thermal noise model, the theoretical level of thermal noise in the indirect-contact ECG was estimated. The results also show that the level of op-amp's intrinsic noise is so small that it can be negligible in comparison with thermal noise of resistors. The relationship between the level of thermal noise and the resistance of the bias resistor was derived, and it is the base for the further study how to choice the optimal resistance for the bias resistor.

• School Mathematics
• /
• v.5 no.2
• /
• pp.167-189
• /
• 2003

Mental arithmetic has recently gained a higher profile in primary school mathematics. The study aims to reflect didactical background of mental arithmetic in number and operations curriculum for primary school mathematics. In order to attain these purposes, the present paper describes the meaning of mathematical literacy and didactical background of mental arithmetic on which have been laid emphasis in relation to mathematical literacy in many countries. Also it shows current suggestions for mental arithmetic instruction in Everyday Mathematics Project in USA, Numeracy Number Project in Great Britain, TAL project based on Realistic Mathematics Education in the Netherlands, and mathe 2000 project in German in order to gain practical ideas for teaching mental arithmetic. Furthermore, it discusses mental strategies of students and didactical models for improving mental arithmetic instruction based on the results of many researches. Under these theoretical foundations, it is analyzed how mental arithmetic is developed in our number and operations curriculum, focused on mental strategies and didactical models. Finally, implications for improving our mental arithmetic instruction are discussed.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.33 no.9
• /
• pp.89-96
• /
• 2005

This paper discusses the daytime atmospheric conditions and the possibility of daytime star detection with the purpose of practical use of the star sensor for daylight navigation. In order to estimate the daytime atmospheric data, we use the standard atmospheric model (LOWTRAN 7), from which atmospheric transmittance and radiance from background sky are calculated. Assuming the star sensor with an optical filter to reduce background radiation, different separation angles between the star sensor and the sun are set up to express the effect of the solar radiation. As considerations of field of view (FOV) of the star sensor, the variation of the sky background radiation and the star density of the detectable star are analyzed. In addition, the integration time to achieve a required signal-to-noise ratio and the number of the radiation-caused electrons of the charge coupled detector(CCD) working as the limit to daylight application of the star sensor are calculated.